Design oversights and security flaws; clumsy coding typos can bring institutions to their knees and unseen bugs can be catastrophic...
Our everyday lives have never been more connected and managed through software. Programs not only power our gadgets but drive our daily routines and ensure the world around us keeps on ticking over.
The services we rely upon exist at the mercy of human error, design oversights and security flaws; clumsy coding typos can bring institutions to their knees and unseen bugs can be catastrophic. Quality assurance testing has never been more important.
Here are five infamous incidents of good software gone bad that underline the value of testing consultancy today.
1. The Sky Is Falling -- Amazon crashes and brings down the cloud
Last year, a storm in the skies above Washington D.C. brought down Amazon's Elastic Compute Cloud, the cloud servers used to host many of the web's most popular sites and services including Netflix, Pintrest and Instagram. While software developers can't be blamed for failing to control extreme weather, the lack of sufficient back-up systems showed up the inherent weakness with their cloud-hosted services
- Don't wait until back-ups are needed before finding out whether they work. Always test and ensure secondary systems are up to scratch.
2. What In The World? -- Apple become fallible with maps
Apple's immaculate public image took a hit following the troubled release of Apple Maps last year. Google's established map service was initially unavailable for the new iOS6 mobile operating system giving Apple app a great opportunity to gain ground on its competitor.
Unfortunately, the application was beset by problems such as inaccurate landmarks, wrong directions and mapping mistakes. Australian police even warned against using the app after a number of people became stranded after Apple Maps sent them 43 miles in the wrong direction.
- Opportunities can't be taken with inaccurate and unreliable software.
3. The Mars Climate Orbiter -- a problem of scale
On September 23, 1999, NASA's Mars Climate Orbiter disintegrated in the Martian atmosphere due to a fatal design flaw in the software designed to control the craft's thrusters.
While one program was making the necessary calculations using metric measurements, another based its workings on imperial. The resulting conflict caused the experiment to burn up on entry costing $125 million and more than a year's worth of work and manpower.
- Communication and fact-checking is vital to prevent costly misunderstandings and mistakes in development.
4. Do The Worm -- Robert Tappen Morris creates a monster
Robert Tappen Morris accidently created the world's first worm in 1989, after infecting over 6,000 computers across America's academic and military networks. Created as a harmless if reckless experiment, Morris made an error when writing the program causing it the copy itself at an accelerated rate, overloading and damaging hard drives.
- Always test programs in a closed, safe environment.
5. Cashing In The Chips -- Intel loses millions through faulty coding
Like Apple, Intel was a respected brand that lost credibility due to a defective product. After a math teacher discovered their much-vaunted Pentium chipset made mistakes due to a flawed division table, the computer giant was forced to replace the bugged chips of anyone who complained, costing Intel a cool $475 million.
- Bugs can be an expensive problem to solve if they're not hunted down before release.
Bio: Roger Trent works in software testing consultancy and has witnessed major corporations fall from the smallest defects.